The Reaction of 8-Nitroquinoline with Thiophenol—Thiophenoxide Ion

ACS2GO © 2019. ← → → ←. loading. To add this web app to the home screen open the browser option menu and tap on Add to homescreen...
0 downloads 0 Views 932KB Size
J O U R N A L OF T H E A M E R I C A N CHEMICAL SOCIETY (Registered in U. S. Patent O5ce)

VOLUME74

(Copyright, 1952, by the American Chemical Society)

JUNE 25, 1952

[CONTRIBUTION FROM THE

DEPARTMENT OF CHEMISTRY

NUMBER 12

OF COLUMBIA UNIVERSITY ]

The Reaction of 8-Nitroquinoline with Thiophenol-Thiophenoxide Ion. An Example of Anionic Substitution1 BY ROBERTC. ELDER FIELD^^ AND ELIZABETH F. CLAFLIN RECEIVED NOVEMBER 23, 1951

The reaction of 8-nitroquinoline with thiophenol and thiophenoxide ion results in reduction of the nitro group and anionic substitution of one or more phenylmercapto groups on the nucleus. Bromine atoms in 5,7-dibromo-8-aminoquinoline are displaced by chlorine during diazotization and deamination in hydrochloric acid.

In a previous communication2 the reaction of 4chloro-8-nitroquinazoline with thiophenoxide ion, during which not only the reactive chlorine but also the nitro group was displaced by phenylmercapto, was reported. Kremer3 obtained preliminary evidence that, $the reaction of 5-bromo-6-methoxy-8nitroquinoline with sodium methyl mercaptide in boiling methyl Cellosolve, the nitro group, in addition to the bromine atom, was displaceable by a methylmercapto or P-methoxyethoxy residue. In order to throw further light on this unexpected lability of a nitro group in the 8-position of quinazoline or quinoline toward anionic reagents we have investigated in detail the reaction of 8-nitroquinoline itself with thiophenol in the presence of thiophenoxide ion. No displacement of the nitro group was observed; rather the reaction took a different course. The reactions of aromatic nitro compounds with organic and alkali metal sulfides and other basic reagents may be divided into three general types. (1) The familiar reduction of nitro groups to amines or intermediate reduction products. ( 2 ) Displacement of suitably activated nitro groups in polynitro compounds by the base during which the basic frag(1) The material presented in this paper is taken from a dissertation submitted by Elizabeth F. Cla5in in partial fulfillnient of the requirements for the degree of Doctor of Philosophy in Columbia University in June, 1951. A complete literature review of anionic substitutions will be found in the dissertation. After this manuscript was submitted, substantially the same general type of reaction with various nitro compounds and mercaptans was reported by H. R . Davis, Jr., J. W. Copenhaver, W. E . Hanford and H. F. Lederle (Abstracts of Papers Presented at the Spring Meeting of the American Chemical Society, Buffalo, N. Y.,March 23-27, 1952, p. 46K). (la) University of Michigan, Ann Arbor, Mich. (2) R . C. Elderfield, T. A. Williamson, W. J. Gender and C. B. Kremer, J . Org. Chem., 11, 405 (1947). (3) C. B. Kremer, unpublished work from these laboratories.

ment may either take the position vacated by the nitro group or appear elsewhere in the molecule. (3) Direct substitution by the base accompanied by some reduction of the nitro groups. The third type of reaction is pertinent to the present work. Bradley and Robinson4 studied the addition of anionic reagents to aromatic systems and classified the ring systems capable of reacting with negatively charged fragments into four types: quinones, nitro compounds, phenols (or amines), such as resorcinol, which can display keto-enol tautomerism, and heterocyclic nitrogen compounds. In particular, they succeeded in adding piperidine in the presence of sodamide to the para position of nitrobenzene, the 4-position of 1-nitronaphthalene, and the 5-position of 8-nitroquinoline (vide infra). Bergstrom, Granara and Ericksons accomplished the addition of diphenylamine to the para position of nitrobenzene in the presence of sodamide in liquid ammonia. The reaction gave the best yield with excess nitrobenzene which they concluded reacted with the displaced hydride ion to form indefinite reduction products. The direct hydroxylation of nitrobenzene by solid potassium hydroxide6 and the Piria reaction for the preparation of amino sulfonic acids from nitro compounds and alkali sulfites and bisulfites’ are other examples of such substitutions. 8-Nitroquinoline fits two of the types in Bradley and Robinson’s clas~ification.~The nitro group is capable of inducing by resonance an electron de(4) w. Bradley and R . Robinson, J . Chem. Soc., 1254 (1932). (5) F. W. Bergstrom, I. M. Granara and V. Erickson, J . Org. Chem, 7 , 98 (1942). (6) A. Wohl, Ber., 82, 3486 (1899); 0. C. Dermer and L. J. Druker, PYOC. Oklahoma Acad. Sci., 18, 55 (1943); P. Hepp, Bcr., 18, 2346 (ISSO), infer alia. (7) R . Piria, A n n . , 78, 31 (1851); W. M. Lauer, M. M. Sprung and C. M. Langkammerer, THIS JOURNAL, 58, 225 (1936), inlcr alia.

2953

ROBERT C. ELDERFIELD AND ELIZABETH F. CLAFLIN

2954

ficiency on the 5- and 7-carbon atoms making them attractive to anions, and the heterocyclic nitrogen exerts an attraction on the ring electrons leaving a partial positive charge on the 2- and 4-carbon atoms. If either set of effects can be extended to the other ring by resonance, the activation will be reinforced thereby increasing the reactivity of the compound toward anionic reagents a t the 2-, 4-,5and 7-positions. When 8-nitroquinoline (I) was heated with sodium thiophenoxide and excess thiophenol in boiling methanol according to the published procedureI2 a mixture of products was obtained. This was separated into its components as shown in Chart 1. CHART1

VOl. 74

has been assigned. Unreacted I (about 25%) was recovered from the combined benzene washings on dilution with hexane. I1 was very unstable and darkened in air or on recrystallization from most solvents in which it turned to a brown gum. It was soluble in hydrochloric acid with decomposition and gave a positive test with Tollens reagent. On this basis combined with elementary analyses, the compound was assigned structure 111. It was reduced with stannous chloride to 8-amino-5-phenylrnercaptoquinoline (V). V was synthesized independently for comparison by reaction of 5-bromo-8-nitroquinoline with one mole of sodium thiophenoxide and subsequent reduction of the nitro group. In order to remove all reasonable doubt as to the structure of VI 8-amino-7-phenylmercptoquinolinewas prepared similarly from 7-chloro-8-nitroquinoline. CHART2

I (1) Na+C&LS--C&SH (2) dil. NaOH

J. Filtrate

+

Br+2HtCOOH

ClrCH*COOH

Br

'y

\i.

Filtrate

Residue

1

!

HCl

IV benzene

$ 7 Filtrate CsH6SSCsHs

Precipitate

Filtrate

Residue

I

I1

I11

1'

The reaction mixture was poured into dilute alkali and the insoluble material was washed thoroughly first with hexane and then with benzene. The insoluble residue was nearly pure 5-phenylmercapto8-quinolylhydroxylamine (11) (50%). From the combined hexane washings the hydrochloride of 8amino-5,7-bisphenylmercaptoquinoline (111) (5%) was obtained. Evaporation of the hexane filtrate from I11 gave an amount of diphenyl disulfide approximately equivalent to the combined yield of I1 and 111. Extraction of the original alkaline filtrate with ether gave a small amount of brilliant orange needles to which the probable structure IV

4

Sn CIz-HCl

I11

If the theoretical considerations discussed above are valid, I11 should possess the structure assigned to it. That this was so was shown by synthesis of I11 as shown in Chart 2. However several points of interest were encountered during the course of this synthesis. 8-Aminoquinoline was brominated by

June 20, 1952

8-NITROQUINOLINE WITH THIOPHBNOL-THIOPHENOXIDE ION

2955

a modification of the procedure of Claus and Setzer.' products of the reaction, no substances lacking the The product of this,reaction had been shown by second nitrogen atom could be isolated. About Claus and Setzer to be identical with the substance 5% of 5-methylmercapto-6-methoxy-8-aminoquinoformed in the Skraup reaction with 2-nitr0-3~5-di- line was isolated. The formation of this indicates bromoaniline. Although repeated recrystallization that methyl mercaptide ion acts as a reducing of VI failed to raise the melting point to that re- agent. ported by Claus and Setzer, the analytical data were The reactions of several other nitro compounds with thiophenoxide ion under the conditions used satisfactory. The behavior of VI when elimination of the amino with 8-nitroquinoline were investigated. 6-Nitrogroup was attempted is interesting. When the quinoline and nitrobenzene did not react. On the and diazotization was camed out in hydrochloric acid basis of Bradley and Robinson's ob~ervation,~ and the crude product was nitrated directly, the since it is electronically analogous to S-nitroquinom.p. of the product could not be raised above 168O, line, m-dinitrobenzene might be expected to add the reported m.p. of 5,7-dichloro-8-nitroquinoline thiophenoxide ion. Under the reaction conditions, (VII) whereas 5,7-dibromo-8-nitroquinoline (VIII) m-dinitrobenzene was reduced to m,m'-dinitromelts a t 198'. The identity of VI1 was established azoxybenzene. No evidence of addition was found. A few observations on the mechanism of the reacby comparison with an authentic sample prepared by the Skraup reaction,$ and by conversion of ma- tion of I with thiophenoxide ion and thiophenol may terial from both sources to 5,7-dimethoxy-8-nitro- be made. The reaction is base catalyzed since only starting material was recovered from the prodq~inoline.~ The only source of chlorine in this exchange reac- ucts of an experiment with neutral thiophenol. It tion was the hydrochloric acid present in the diazo- is also likely, from the observation of Bradley and tization. Evidently there was reaction either with Robinson' that the formation of the thio ether is an the conjugate acid of VI or with the diazonium salt. anionic reaction. If addition of an anion to an aroAt first sight the former seemed to be more likely. matic system occurs, a hydride ion must either be Lauer'O noted that 5-bromo-6-methoxy-8-acetamino- discharged or accounted for by an inter- or intramoquinoline is converted to 7-bromo-6-methoxy-8- lecular reduction. Discharge of hydrogen gas from acetaminoquinoline in 85% yield when it is boiled this reaction mixture was precluded by the fact that with concentrated hydrochloric acid. A cationic the reaction would then demand the formation of a exchange mechanism was postulated. However larger amount of diphenyl disulfide than the one when 5,7-dibromo-8-aminoquinolinewas warmed equivalent actually found and by the fact that no 5for two days on the steam-bath with hydrochloric phenylmercapto-8-nitroquinolinewas isolated from acid-conditions more drastic than those obtained the products. Treatment of 5-phenylmercapto-8during the diazotization reaction-little, if any, ex- nitroquinoline with thiophenol-thiophenoxide ion change occurred. Therefore, the exchange must under the usual conditions of the reaction resulted have taken place during the diazotization-deamina- in the recovery of 70-90% of unchanged starting material, so that it seems unlikely that this subtion step. Diazotization of VI in other acids was less satis- stance would have been overlooked if it were presfactory because of insolubility or formation of a ent in significant amounts in the reaction products. by-product, m.p. 134O, which was not investigated 5-Phenylmercapto-8-nitroquinolinewould also be further. Diazotization directly in 50% hypophos- expected among the reaction products. From these phorous acid gave about 40% of 5,7-dibromoquino- facts it may be concluded that the addition of one mole of thiophenoxide ion to I is accompanied by an line (IX) which was nitrated to VIII. 5,7-Dichloro-8-aminoquinolinewas prepared by intramolecular hydrogen transfer to the nitro (or direct chlorination of 8-aminoquinoline and diazo- nitroso) group, which could be accomplished by a shift of two electrons into the aromatic nitro systized, deaminated and nitrated to yield VII. Either VI1 or VI11 on reaction with sodium thio- tem, and an exchange of protons with the solvent phenoxide gave 5,7-bis-phenylmercapto-8-nitro-(thiophenol), (see chart). A similar scheme may be quinoline (X) which on reduction gave 111,identical written for 8-nitroquinoline. It is not likely that thiophenoxide ion adds to 8with the substance obtained from the reaction of I quinolylhydroxylamine to give 11. However, the with thiophenoxide ion and thiophenol. During the study of 11, it was reduced with Ra- possibility that I1 is an intermediate in the formaney nickel containing hydrogen according to the tion of I11 is not out of the question. When 8method of Mozingoll in order to remove the phen- quinolylhydroxylamine was treated with thiopheylmercapto group. The product was 8-amino-1,2- nol-thiophenoxide ion under the usual conditions about 10% of I11 was obtained. 3,4-tetrahydroquinoline. We believe that the formation of I1 is preceded by The action of methoxide and 8-methoxyethoxide ions on 8-nitroquinoline was investigated. No reduction of I to 8-nitrosoquinoline by thiophenol, evidence of displacement was noted. The reaction rather than by addition of thiophenoxide ion to 8reported by Kremers was repeated and, although a nitroquinoline and subsequent reduction. Although definite test for nitrite was given by the inorganic i t was not possible to demonstrate this experimentally because of our failure to prepare 8-nitroso( 8 ) A. Claus and E. Setzer, J . prakf. Chcm., [2] 68, 401 (1896). quinoline, certain observations in the literature sup(9) R. C. Elderfield and G . L. Krueger. J . Org. Chcm., 17, 858 (1952). port such a thesis. (10) Private communicatiou from Dr. W. M. Lauer. The nitroso group has been reported to be para (11) R. Mozingo, D. E. Wolf, S. A. Harris end K. Folk,rs. THIS directing in aromatic substitution reactions such as JOURNAL. 66, 1015 (1943).

-

and what was described as 5-piperidino-8-nitroquino h e . Neither of these substances was identical with the product of their reaction. However, in NO their synthesis of 5-piperidino-g-nitroquinoline, they proceeded from a chloronitroquinoline (m.p. 185') to which the structure of 5-chloro-8-nitroquinoline had been erroneously assigned.ls It was subsequently I 'I /I shownlC that this is the 7-chloro K --OS N=O s=-0 compound and that the 5-chloro CBH6SH isomer melts a t 138'. The 2and 7-isomers had, therefore, been C0H;SWII CSI I,,S"II eliminated. We have synthesized C6H5St 5-piperidino-8-nitroquinoline from t----+------CeHsS'l + C ~ H ~ S C5-bromo-&nitroquinoline and ob\ N/ \i$ \N/ tained a substance of the same map. as that obtained by Bradley NHOH I1 it C&SH HN-OH X-OH and Robinson by addition of @ piperidine to 8-nitroquinoline. halogenation and nitration.12 From the rule of Attack by thiophenoxide ion as here reported is Harnmick and I l l i n g ~ o r t hthe ' ~ nitroso group would therefore consistent with the observations of be expected to be meta directing like the nitro Bradley and Robinson, group. However, nitrosobenzene can be expressed Experimental l7 by resonance forms in which the ortho and para carbon atoms have either increased or decreased Reaction of 8-Nitroquinoline (I) with Thiophenol-Thioelectron density. phenoxide Ion.-I was recrystallized from 3 :1ethanol-ether I

+ 2CaHrSH

+ CeH5SSCeHs+ HzO

c6Ht() '

4

07

+

XI

The strong activating effect of the nitroso group toward anionic reagents is evident from the ease with which substituents can be displaced from the para position by bases. Thus in the reaction of p-bromonitrosophenol with silver nitrate, the p-nitroso group has a labilizing effect far greater than that of two nitro groups in the 2- and 4-positions and comparable to, if not greater than, that of three nitro groups in picryl chloride or bromide.I4 Thus, toward anionic substitution in nitrosobenzene, the contribution of resonance form XI would be expected to be much greater than that of the analogous form of nitrobenzene, and there is no reason why a similar state of affairs should not exist with the 8-substituted quinolines. The anionic substitutions of Bradley and R ~ b i n s o nof, ~Bergstrom and c o - ~ o r k e r sand , ~ the hydroxylation of nitrobenzene6 were all accomplished under much more drastic conditions than the present or known substitutions on nitroso compounds and were not accompanied by reduction. The Piria reaction proceeds under milder conditions and is accompanied by reduction. As previously mentioned, Bradley and Robinson4 successfully added piperidine to 8-nitroquinoline in the presence of sodamide to form a piperidino-8nitroquinoline. The position of the piperidino group was not determined. Based on the prediction that this would occupy the 2-, 4, 5- or 7-position, they prepared 2-piperidino-8-nitroquinoline, (12) C. K. Ingold, J . Chem. Soc., la?, 513 (1925).

(13) D, I,. Hammick a n d W. S. Illingworth, ihid.. 2358 ( I ! W ) . K. J. W. Le Pevre, i b i d . , 810 (1931).

(14)

to a constant m.p. of 86-89'. A solution of 5 g. of I, 0.6 g. (1 equiv.) of sodium and 20 g. (6 equiv.) of thiophenol in 100 ml. of anhydrous methanol was refluxed for two hours under nitrogen, and then poured into one liter of water and cracked ice containing 6 g. of sodium hydroxide. After 20 minutes the precipitate had coagulated sufficiently to be filtered. The precipitate was sucked as dry as possible yielding 14.2 g. of a gummy yellow solid. It was washed twice with hexane by decantation and again filtered. The hexane washings were worked up as described below. The hexane insoluble material (5.8 g.) was washed twice with cold benzene by decantation, the washings were filtered, diluted with hexane and allowed to stand. The b e y zene-insoluble material (11) (3.8 g . , 50%), m.p. 96-100 , was repeatedly recrystallized from dry benzene to yield a microcrystalline substance, m.p. 99.8-100.2° (dec.). Anal. Calcd. for ClbH12N2OS: C, 67.2; H, 4.5; S , 10.5; S, 11.9. Found: C,67.,5; H, 4.5; S, 10.4; S, 11.8. The compound decomposed very readily to a red or dark brown gum if it was allowed to stand in air for more than a few minutes before the washing with benzene was completed. It also decomposed on heating or on recrystallization from alcohol or wet benzene. The thoroughly washed compound was more stable. It is soluble in acid with decomposition and insoluble in base. The Tollens test, i n alcoholic solution, was positive. From this observation taken together with its instability and the isolation of one equivalent of diphenyl disulfide from the hexane-soluble fraction, the structure of a hydroxylamine was assigned to the substance. The yield varied widely. It was lowered by use of less than 6 equiv. of thiophenol, by use of more than 1.4 equiv. of sodium, or by use of I of m.p. below 86'. The hexane washings of the crude reaction product were combined and saturated with dry hydrogen chloride. The precipitate was shaken in a separatory funnel with ether and dilute ammonia solution. The dried ether solution on evaporation left a yellow oil. Two recrystallizations from hexane gave 0.35 g. of yellow solid, m.p. 80-88'. The remaining two-thirds of the oil could not be crystallized. Fur(15) A. Claw and K . Junghanns, J . p v n k l . L'hcm., [ 2 ] 48, 253

(Ism), (10) (a) L. Bradford, T. J. Eliot a n d F. M . Rowe, J . Chem. Sod., 437 (1947); (b) C. C. Price and D. B. Guthrie, THISJOURNAL, 68, 1592 (1040). (17) hiicroanalyses by Clark Microanalytical Laboratory, Urbana, Ill., and Schwarzkopf Microanalytical Laboratory, Middle Village, L. I., N. Y. A l l m.ps. arc unrorrectrd rxrept those reported t o tenth::

of a d r p r r .

June 20, 1952

8-NZTROQUINOLINE WlTH

THIOPHENOL-THIOPHENOXIDE ION

2957

ther recrystallization of the solid material from alcohol or ml. of methyl Cellosolve during 15 minutes. The mixture was refluxed for two hours and allowed to stand overnight, hexane gave fine, greenish-white needles, m.p. 97.7-98.4'. The substance showed a strong blue-white fluorescence in After removing the solvent under reduced pressure, the resiultraviolet light and gave a bright green color in concd. sul- due was taken up in boiling absolute alcohol and filtered from furic acid when a drop of nitric acid was added. Analytical inorganic salts. The salts gave a strong test for nitrite ion with starch-potassium iodide test paper and a rough quandata corresponded to those demanded by 111. Anal. Calcd. for CzlHlaN&: C, 70.0; H, 4.4; S, 17.8. titative estimate showed the ratio of bromide to nitrite to be about 4: 1. Nevertheless no quinoline derivatives in which Found: C, 70.0; H, 4.6; S, 17.3. The acetyl derivative of I11 was prepared by allowing it the nitrogen in the 8-position was missing could be isolated. The alcoholic filtrate pn cooling deposited 8.2 g. of tan to stand overnight in acetic anhydride and collecting the crystals, m.p. 124-129 Recrystallization from alcohol precipitate. Aofter recrystallization from alcohol it melted gave 4.8 g. of brownish-gold plates, m.p. 130-134'; reat 184.1-184.3 ported m.p. for 5-methylmercapto-6-methoxy-8-nitroquinoA w l . Calcd. for C23H18N20S2: C, 68.6;H, 4.5. Found: line, 136O.3 The crude yield was 18%. C, 68.8; . H,. 4.8. The mother liquor from the above compound was diluted The picrate of I11 came down as dark purple prisms from with an equal volume of ether and filtered from black, hyalcohol, m.p. 134-136'. Several recrystallizations from groscopic, amorphous material. From the filtrate about 1 ethyl acetate gave thick orange needles, m.p. 140.5-141.5'. g. of yellow needles, m.p. 171-172' after recrystallization Intermediate recrvstallizations gave both types of crystals. from dry benzene, was obtained; reported m.p. for 5Anal. Calcd. -for C ~ P H ~ ~ N ~C, ? ~-55.0; ~ S ~ : H, 3.2. methylmercapto-6-methoxy-8-aminoquinoline, 166-167O.3 The substance is readily absorbed on Norite and can be Found: C, 54.7; H, 3.6. Evaporation of the hexane filtrate from the hydrochloride eluted with benzene. Anal. Calcd. for C1LHltN20S: C, 60.0; H, 5.5; S, 14.7. of I11 left 4-6 g. of diphenyl disulfide, m.p. 57-60". The benzene wash liquors of the crude reaction product on Found: C, 59.9; H, 5.6; S, 14.9. dilution with hexane furnished 0.8 g. of unreacted I. For comparison, a solution of 0.25g. of 5-methylmercaptoIn an experiment in which 20 g. of I was added t o 400 ml. 6-methoxy-&nitroquinoline, m.p. 135-137°,ain 2.5 ml. of of anhydrous methanol in which 40 g. of thiophenol (3 concd. hydrochloric acid was added dropwise with vigorous equiv.) and 5 g. of sodium (2equiv.) had been dissolved, the stirring to a solution of 1 g. of stannous chloride in 8 ml. of mixture suddenly turned dark after two hours' heating. concd. hydrochloric acid at 5 ' . The solution was stirred After pouring into ice and sodium hydroxide solution, the one hour in the ice-bath and for an additional hour at room dark brown moist precipitate (39g.) was washed three times temperature. After standing overnight it was diluted with with hexane and the residue (19 9.) was washed with ben- 200 ml. of water and the precipitate dissolved. After coolzene in which most of it dissolved. Treatment of the hex- ing to , ' 5 concd. sodium hydroxide solution was added at ane soluble portion with hydrogen chloride gave 3.9g. of the 5-10" until the tin salts dissolved. From the solution, 0.20 hydrochloride of 111. g. of yellow micro plates, m.p. 170-171' after recrystallizaThe basic filtrate from the crude product was extracted tion from alcohol, were obtained. The mixture m.p. with with ether. Concentration of the dried ether solution gave the substance obtained above was 170-171'. 1 g. of brilliant orange needles, m.p. 187-188' after recrysThe acetyl derivative formed buff plates, m.p. 104-105', tallization from acetone. The substance turned a brilliant from dilute alcohol. red-violet in acid. In the diphenylamine test for nitroso Anal. Calcd. for CI~HI&OZS: C, 59.5; H , 5.3. Found: compounds the solution first turned violet which faded to a C, 59.4;H , 5.2. pale blue-green on standing overnight. With Liebermann Reduction of I1 with Stannous Chloride (V).-The usual nitroso reagent it gave a grass-green color. When it was coupled with aniline in acetic acid solution, a deep crimson stannous chloride reduction procedure gave large amounts solution resulted. Analytical data agreed with those de- of tar. The following was satisfactory. A suspension of finely ground I1 (1.5 9.) in 25 ml. of water manded by IV. was added dropwise to a well cooled, stirred solution of 3 g. Anal. Calcd. for CnHlaN20S2: C, 67.4; H, 3.7; N, of stannous chloride in 28 ml. of 'concd. hydrochloric acid. 7.5; S, 17.1. Found: C, 67.3; H , 3.7; N, 7.4; S, 17.5. After addition of 10 ml. of methanol, the mixture was alWhen the reaction was run without a solvent, no I1 was lowed to come to room temperature. A fine orange precipifound. The product was almost exclusively 111. tate formed. The solution was cooled below 10' and 20 g. When I was heated with thiophenol in methanol in the of sodium hydroxide in the form of a 40% solution was absence of sodium no reaction occurred. added with stirring and cooling a t such a rate that the temReduction of I1 with Raney Nickel.-A suspension of 7-10 perature did not exceed 10'. Two more portions of 20 g. of g. of Raney nickel prepared according to Mozingo, et al.," sodium hydroxide were added and the solution was exin a solution of 0.76 g. of I1 in 15 ml. of absolute ethanol was tracted with ether. The dried ether extract gave 1.3g. of a refluxed for 45 minutes. After filtration from the catalyst yellow-brown oil. This was taken up in hexane, treated and evaporation of the solvent a dark brown viscous oil re- with charcoal, and cooled in a solid carbon dioxide-bath mained. A picrate prepared from a filtered absolute alcoholic yielding 0.5 g. of crystalline material, m.p. 76-82'. Resolution of the oil melted at 183-187' and did not depress the peated recrystallization from hexane, at first from the carbon m.p. (182-184') of a known sample of the picrate of 8-amino- dioxide-bath, gave flat yellow needles, m.p. 88.9-90.1"; 1,2,3,4-tetrahydroquinoline.l8 yield 30-35%. Further confirmation of the identity of the %aminoAnal. Calcd. for CllH12N&: C, 71.4; H , 4.8; N, 11.1; 1,2,3,4-tetrahydroquinolinewas provided by its conversion S, 12.7. Found: C, 71.3; H.4.8; K, 10.9;5, 12.3. to 2,2-dimethyl-1,2,5,6-tetrahydro-4-imidazo[ij]quinoline, - The acetyl derivative formed white.needles from alcohol, m.p. 139-141', when it was warmed with acetone.le m.p. 161-162', and the mixture m.p. with a known sample Anal. Calcd. for C12H16NZ: C, 76.6; H, 8.5. Found: was not depressed. C, 76.0; H, 8.3. The p-toluenesulfonyl derivative formed pale greenish On thermal decomposition this was converted to 2-methyl- plates from alcohol, m.p. 155-156', and the mixture m,p. 5,6-dihydro-4-imidazo[ij]quinoline,m.p. 127'; reported with a known sample was not depressed. m.p. 128'.'* 5-Phenylmercapto-8-nitroquinolh1e.-To a hot solution Reaction of 5-Bromo-6-methoxy-8-nitroquinolinewith of 11.5 g. of 5-bromo-&nitroquinoline (m.p. 147-150') in Sodium Methyl Mercaptide.-The procedure was a modifi- 450 ml. of alcohol was added a solution of 1 g. of sodium and cation of that of KremerJ and of Gilman, el a1.a To a solu- 5 g. of thiophenol in 25 ml. of methanol. After refluxing tion of 29 g. of 5-bromo-6-methoxy-8-nitroquinoline(m.p. for 1.5 hours, the mixture was evaporated to 200 ml. and 203-206') in 250 ml. of methyl Cellosolve was added a solu- poured into 800 ml. of yater. The precipitate (12.4 g., tion of 16 g. of sodium and 22 g. of methyl mercaptan in 180 97%) melted at 140-143 Recrystallization from alcohol gave yellow prisms, m.p. 142.3-143.4'. (18) S. J. Hazlewood, G. K. Hughes and F. Lions, J . PYOC. Roy. SOC. Anal. Calcd. for C15H10N202S: C, 63.8; H, 3.6; S, N . S. Wales, 71, 467 (1937-1938). 11.4. Found: C, 63.8; H, 3.5; S, 11.4. (19) Cf. H.J. Barber and W. R . Wragg, J . Chem. Soc., 610 (1946); The picrate, m.p. 138.2-13!?.1', formed orange prisms R.C.ElderEeld and F. J. Kreysa, THIS JOURNAL, 70, 4 4 (1948). from alcohol. (20) H . Gilman, el at., ibid., 68, 1578 (1946).

.

.

~~

.

2058

RORERTc. %DRRFIELD

AND E1,IZARBTH

Anal. Calcd. for C Z ~ H ~ J N ~ O N,~ S13.7. : Found: N, 13.6. 5-Phenylmercapto-8-aminoquinoline (V).-The above substance was reduced with stannous chloride and hydrochloric acid.ll When the reaction mixture was diluted the tin complex did not dissolve. After filtering it and drying, the yield of tin complex was 75%. A suspension of 3.6 g. of the tin complex in ether and water in a separatory funnel was made alkaline with 10 g. of sodium hydroxide. From the ether 2.2 g. of V was obtained as long, yellow, iridescent needles, m.p. 85-87'. Recrystallization from hexane raised the m.p. to 90.4-90.7'. The mixture m.p. with V obtained as above was not depressed, nor were the mixture m.p.'s of the following derivatives. Anal. Calcd. for C ~ ~ H I Z NC, ~ S71.4; : H, 4.8; N, 11.1; S. 12.7. Found: C, 71.4; H , 5.0; N, 11.3; S, 12.2. The p-toluenesdfonyl derivative formed pale greenish plates, m.p. 159.1-160.0' from alcohol. Anal. Calcd. for C=HI~N~O&:C, 65.0; H, 4.5; S, 15.8. Found: C, 65.2; H, 4.5; S. 15.6. The acetyl derivative formed white needles, n1.p. 162.5163.0', from 70% alcohol. Anal. Calcd. for C17H1,NzOS: C, 69.4; H, 4.8. Found: C, 69.3; H, 4.9. 7-Phenylmercapto-8-nitroquinoline.-This was prepared as was the 5-derivative except that it was necessary to reflux the mixture six hours. The crude yield was 97%. Recrystallization from alcohol gave yellow plates, m p. 98.899.6'. Anal. Found: C , 63.5; H, 3.6; S,11.3. The picrate formed short, deep yellow needles, m.p. 144.6-145.5', from alcohol. Anal. Found: N, 13.5. 7-Phenylmereapto-8-aminoquinoline,rn .p. 68.2-69.4' after recrystallization from hexane, was prepared as was the .?-derivative. Anal. Found: C, 71.6; H, 4.5; N, 10.9. The p-toluenesulfonyl derivative, m.p. 197.0-197.8°, formed white plates from alcohol. Anal. Found: C, 65.1; H, 4.4. The acetyl derivative, m.p. 127.0-127.9", formed hygroscopic white needles from 50% alcohol. Anal. Found: C, 68.1; H, 5.2. 5,7-Dibromo-8-aminoquinoline (VI).-The method of Claus and Setzerg was used with certain necessary modifications. To a solution of 3.6 g. of 8-aminoquinoline in 30 ml. of glacial acetic acid and 8 ml. of water cooled to -15' was added dropwise with stirring a solution of 8 g. of bromine in 25 ml. of 80% acetic acid. After a portion of the bromine had been added, the mixture set to a thick mush. Filtration gave 1.25 g. of yellowish-gray powder, m.p. 114-117'. This was free VI. The filtrate was returned to the flask, cooled to -12', and the remaining bromine solution was added. This gave a precipitate of 6.2 g. of the hydrobromide of VI. Dilution of the filtrate gave an additional 0.4 g. of free VI, m.p. 113-117'. The f i s t two fractions were suspended in 50 ml. of concd. hydrochloric acid and the mixture was diluted to 2 1. yielding 5.3 g. of light gray powder, m.p. 116-118'. On recrystallization from hexane it formed long, flat, pale yellow needles, m.p. 120.3-120.6 . We were unable to raise the m.p. to 127' as reported by Claus and Setzer.s Anal. Calcd. for C911bBr2N2: C, 35.8; H, 2.0; Br, 52.9. Found: C,36.3; H, 1.7; Br, 53.4. The acetyl derivative formed white needles, m.p. 212.6213.6". dilute alcohol. ~ . _ from . Anal. Calcd. for CllHsBr~NzO:C, 38.4; H. 2.3. Found: C, 38.6; H, 2.5. Diazotization and Deamination of VI.-A suspension of 5.7 g. of VI in 75 ml. of concd. hydrochloric acid was diazotized a t 0' with the calculated amount of sodium nitrite. After addition of 25 ml. of 50% hypophosphorous acid the solution was left for one day in the ice-box and for an additional day at room temperature and then poured into 1.5 1. of water. After filtration from some 1.5 g. of pinkish solid, m.p. 133-135', which was not investigated further, (21)

K. C. Elderfield, et al., THISJOURNAL, 68, 1680 (1940)

F. CLAFLJN

Vol. 74

the filtrate was made basic and 4.2 g. of white solid separated. This was recrystallized repeatedly from alcohol yielding white needles, m.p. 114.8-115.2'. Claus and Setzera report m.p. 112' for 5,7dibromoquinoline, and rn.p. 116" for 5,7-dichloroquinoline. Anal. Calcd. for CoHsBr2N: C, 37.7; H, 1.7. Calcd. for C H d 3 N : C, 54.6; H, 2.5. Found: C, 55.5; H, 3.1. Although the analytical data are not good, they suffice to show that exchange of bromine for chlorine had occurred. The best yield was 60%. Subsequent runs, worked up by subliming the material precipitated by base, gave a product melting a t 109-111' which was nitrated directly since the impurities were oxidized during nitration. In view of the above result, deamination of VI in acids other than hydrochloric was investigated. A solution of 0.50 g. of VI in 20 ml. of 50% hypophorous acid, 4 ml. of concd. sulfuric acid and 6 cc. of water was diazotized with 0.12 g. of sodium nitrite and allowed to stand for two clays in the ice-box and one day a t room temperature. The crude product isolated as above was sublimed in DUCUO yielding 0.19 g. of impure IX, m.p. 96-98', which, however, was satisfactory for nitration. Diazotization in sulfuric acid gave primarily the by-product, m.p. 135'. Nitration of 5,7-Dichloroquinoline (VII) .-The crude Droduct of the deamination of VI. m.D. 102-109' f0.6 E.) &as added to a mixture of 3 ml. of con'cd. sulfuric acid &d 3 ml. of fuming nitric acid cooled in an ice-bath. After standing several hours a t room temperature, the mixture was heated on the steam-bath. The course of the reaction was followed by diluting a drop of the mixture with water and noting the m.p. of the precipitate. After eight hours a t room temperature this was 150-161' and, after three hours on the steam-bath, it was constant a t 163-168'. After pouring the mixture into 100 ml. of water, the precipitate was recrystallized from alcohol. The m.p. and mixture m.p. with a known sample of V I P was 164-167'. Further recrystallization from alcohol raised the m.p. to 168.1169.2". Clam and Amrne1burg"report m.p. 168.5' for VII. Anal. Calcd. for GH&NzOZ: C, 44.5; H , 1.6. Calcd. for GH4Br2NYO2: C, 32.6; H, 1.2. Found: C, 44.9; H, 1.8.

5,7-Diphenylmercapto-&nitroquinoline.-A

solution

of

0.46 g. of VI1 in 30 ml. of anhydrous methanol was added to

a solution of 0.11 g. of sodium and 0.42 g. of thiophenol in 5 ml. of methanol. After refluxing the mixture for 45 minutes, the heavy yellow precipitate, m.p. 184-188', was collected. A further amount was obtained on dilution of the mother liquor. The yield was 90%; Recrystallization from alcohol raised the m.p. to 190-191 Anal. Calcd. for C ~ I H ~ ~ N ~ OC,& :64.6; H, 3.6; S, 16.4. Found: C, 64.5; H , 3.9; S, 16.2. 5,7-Dichloro-8-aminoquinoline .-Chlorine was passed into a solution of 0.5 g. of 8-arninoquinoline in 2.5 ml. of acetic acid until the weight increased 0.5 g. The red precipitate (0.5 g.) was shaken with ether and dilute sodium hydroxide solution. From the ether layer, after drying over anhydrous magnesium sulfate, 0.23 g . (31%) of crude 5,7-dichloro-&aminoquinoline, m.p. 113-1 16', was obtained. Repeated recrystallization from hexane raised the m.p. to 121.3-121.6'; reported m.p. 124OZ2and l19-120'.'6~b) Anal. Calcd. for GHoCIZN~: C, 50.7; 13, 2.8; Cl, 33.3. Found: C, 50.9; H, 2.9; C1,32.8. The chlorination proceeded better at or above room temperature than in an ice-bath. At low temperatures there was considerable blackening. At high temperatures the rate of chlorination exceeded the rate of decomposition. 5.7-Diphenylmercapto-8-aminoquinoline (III).-A hot suspension of 0.3 g. of X in 25 ml. of concd. hydrochloric acid was added dropwise with Vigorous stirring to a solution of 1.2 g. of stannous chloride in 4 ml. of concd. hydrochloric acid cooled to 3-5'. The mixture was allowed to stand for three days and diluted with 200 ml. of water. The tin complex (0.6 9.) was filtered off and shaken with ether and dilute sodium hydroxide solution. From the ether layer 0.16 g. (58%) of 111, m.p. 97.5-98.1' after recrystallization from hexane was obtained. The mixture m.p. with I11 isolated from the reaction of I and thiophenol was undepressed. Anal. Found: C , 70.2; H, 4.7; N, 7.7; S, 17.6. The picrate, m.p. 134-137', and the acetyl derivative,

.

(22) A. Claiiv and A. Ammrlburg. J . prakl. Chcm., [ 2 ] 61, 418 (189.5).

June 20, 1952

2-(3’,4’,5’-TRlMETHOXYBENZOYL)-4,5-METHYLENEDIOXYSTYRENE

2950

m.p. 179-180°, also did not depress the m.p.’s of the corresponding derivatives of I11 obtained as above.

dium hydroxide. The gummy precipitate dissolved almost completely in hexane. The hexane solution yielded a mixReaction of m-Dinitrobenzene with Thiophenol and Thio- ture of 0.9 g. of white needles and a yellow oil which was phenoxide Ion.-The reaction was carried out as was done again taken up in hexane and treated with hydrogen chlowith 8-nitroquinoline. From 2 g. of m-dinitrobenzene 1.05 ride. The precipitated red hydrochloride (0.39 g.) was g. of fluffy white needles separated when the reaction mix- filtered. Evaporation of the hexane gave 0.6 g. of diphenyl ture was poured into 250 ml. of water containing 2 g. of disulfide. The hydrochloride was shaken with dilute amsodium hydroxide. Recrystallization from alcohol or ben- monia and ether. Evaporation of the dried ether extract zene gave m,m’-dinitroazoxybenzene, m.p. 147.9-148.5”; left 0.26 g. of a yellow gum which gave a small amount of reDorted m.D. 143O.23 11, m.p. &8Q0 after crystallization from hexane. The Anal. Calcd. for CIgH8NIOI:’ C, 50.0; H, 2.8. Found: mixture m.p.’s of this with XI and of the acetyl derivatives were not depressed. C,50.0; H,2.9. 5-Piperidino-8-nitroquinoline.-5-Bromo-S-nitroquinoline From the alkaline mother liquor, 3.3 g. of diphenyl disul- (1 g.) was heated for four hours in 10 ml. of piperidine. The fide was obtained. yellow gum which separated when the mixture was Reaction of 8-Quinolylhydroxgl~eand Thiophenol- heavy Thiophenoxide Ion.-A solution of 0.6 g. of S-quinolylhy- poured into water slowly crystallized; yield 1 g. After droxylamineS4in 6 ml. of anhydrous methanol was added to recrystallization from alcohol, yellow prisms, m.p. 131.2132.3”,were obtained. The product obtained by Bradley 10 ml. of methanol containing 0.13 g. of sodium, and 3 g. of thiophenol was heated under reflux for 1.5 hours and and Robinson‘ from 8-nitroquinoline and piperidine in the poured into 50 ml. of water and ice containing 0.8 g. of so- presence of sodamide is reported as melting at 131.5-132.5”. Anal. Calcd. for ClrH1bN202: C,65.4;H, 5.9; N,16.3. (23) C. A. Lobry de Bruyn, Rae. frao. chim., 20, 115 (1901). Found: C, 65.4; H, 5.6; N, 16.0. 62, 1648 (24) L. F. Fieser and E. B. Hershberg, THISJOURNAL, (1940).

[CONTRIBUTION FROM THE CHEMISTRY

NEWYORK27,N. Y.

DEPARTMENT OF BOSTON UNIVERSITY ]

The Single-Stage Conversion of 1-(3’,4’,5’-Trhethoxyphenyl)-6,7-methylenedioxy3,4-dihydroisoquinolineto 2-(3’,4’,5’-Trimethoxybenzoyl)-4,5-methylenedioxystyrene1~2 BY WALTERJ. GENSLERAND CARLOSM. SAMOUR RECEIVED NOVEMBER 12.1951 Evidence is presented for the reaction path in the formation of 2-(3’,4’,6’-trimethoxybenzoyl)-4,5-methylenedioxystyrene by treatment of 1-(3’,4’,5’-trimethoxyphenyl)-6,7-methylenedioxy-3,4-dihydroisoquinoline with methyl sulfate and alkali.

We have recently described a method of converting 1-(3’,4’,5’-trimethoxyphenyl)-6,7-methylenedioxy-3,4-dihydroisoquinoline(I) directly to 2-(3’,4’,5’ trimethoxybenzoyl) - 4,5- methylenedioxystyrene (VII) by treating the dihydroisoquinolinewith methyl sulfate and alkali.a This transformation had suggested itself originally on considering a sequence of separate reactions dealing with hydrastinine and related compounds.* In terms of the present compounds, the sequence is given in formulations I to VII. Now evidence is adduced for this as the path actually followed in the one-stage process. The mode of attack consisted in the preparation of intermediates and test of their behavior with methyl sulfate and with alkali. That alkali alone was ineffective in the conversion of the dihydroisoquinoline (I) to the vinyl ketone (VII) was shown by the resistance of the former compound to the action of hot sodium hydroxide solution. The dihydroisoquinoline was recovered unchanged after long boiling with 20% aqueous alkali. The quaternary dihydroisoquinolinium metho-

-

(1) This work was supported by American Cancer Society Grant-inAid No. CBC-6 as recommended by the Committee on Growth of the National Research Council. (2) A summary of the material in this paper was presented in Boston, Mass., on April 3, 1951, before the Division of Organic Chemistry of the American Chemical Society. (3) W. J. Gender and C. M. Samour, THISJOURNAL, 72, 3318 (1950); 71, 5555 (issi). (4) (a) H. Becker, 6; ai., Ann., 891, 299, 321 (1913); M. Freund, Bcr., 22,2329 (1889); (b) W.Roser, Ann., 249, 156 (1888): M. Freund and F. Becker, Be?., SS, 1521 (1903); H.Decker and P. Beeka, Ann., 894,328 (1913); (c) W.H.Perkin, Jr., J . Chew. Soc., 109,815 (1916).

sulfate (I1 methosulfate) was prepared by allowing the dihydroisoquinoline (I) to react with methyl sulfate in inert solvent; the iodide of I1 was formed using methyl iodide. The corresponding chloride was obtained by reaction of the carbinolamine (111) with hydrochloric acid. Treatment of the quaternary dihydroisoquinolinium salts (11) with aqueous alkali produced the carbinolamine (111) in a virtually instantaneous reaction; but the latter compound remained unchanged even after long exposure to alkali. On the other hand, when the carbinolamine (111) was warmed with alkali and methyl sulfate, the vinyl ketone (VII) was produced in 84% yield. No attempt was made to isolate the two forms I11 and I V of the carbinolamine. Judging from the absence of any carbonyl peak in the infrared absorption curve of the solid (as a mull with mineral oil) structure I11 is preferred to I V for the crystalline material .s In experiments directed to the preparation of the dimethylamino ketone (V) it was found that the action of methyl sulfate in toluene on the carbinolamine produced the quaternary methosulfate of compound VI as the only isolated material. Methyl iodide in chloroform solution transformed the carbinolamine into a mixture from which the iodides of quaternary salts I1 and VI could be isolated; but (5) This is not necessarily the case for solvlions of the carbinolamine. A preliminary study of the infrared abaorption in lolventa such a s chloroform, dihydrofuran and dioxane indicates that the relativeamount of the carbinolamine (III) and of the monomethylamino ketone (1V)the latter form being detected by an absorption peak at 6.0~-depends on the nature of the solvent.